Photometry and Radiometry

advertisement
Photometry and Radiometry
_____________________________________________________________________________________
A Tour Guide for Computer Graphics Enthusiasts
by
Ian Ashdown, P. Eng., LC, FIES
President
byHeart Consultants Limited
This tutorial has been adapted from the book Radiosity: A Programmer’s Perspective by Ian
Ashdown, © October 2002 byHeart Consultants Limited. (Originally published by John Wiley &
Sons in 1994.) Visit http://www.helios32.com to order the book on CD-ROM.
1.
Introduction
What is light? If you read an introductory book on computer graphics, you may learn that light
has “intensity” or “brightness” ... and nothing more! If we are truly interested in creating
physically realistic images, we shall have to do better than this.
What is light? Aristotle (384-322 B.C.) believed that light consists of “corpuscles” that emanated
from the eye to illuminate the world. Today we favor the theory of quantum mechanics (e.g.,
Hecht and Zajac 1987) or perhaps the possibility that light may be vibrations in the fifth
dimension of ten-dimensional hyperspace (e.g., Kaku 1994). Even so, the true nature of light
remains a mystery. It is perhaps appropriate that the preeminent dictionary of the English
language describes light as:
light, n. 1. The natural agent that stimulates the sense of sight. 2. Medium or condition of
space in which sight is possible.
The Concise Oxford English Dictionary
Fifth Edition, 1964
Whatever it may be, our interest in light is much more parochial. We simply want to understand
something about how light is measured and the units in which these measurements are expressed.
Once we understand these concepts, we will have a much better idea of what it is we are trying to
model in our virtual worlds.
2.
What is Light?
Light is electromagnetic radiation. What we see as visible light is only a tiny fraction of the
electromagnetic spectrum, extending from very-low-frequency radio waves through microwaves,
infrared, visible and ultraviolet light to x-rays and ultraenergetic gamma rays. Our eyes respond
to visible light; detecting the rest of the spectrum requires an arsenal of scientific instruments
ranging from radio receivers to scintillation counters.
A-1
A rigorous and exact description of electromagnetic radiation and its behavior requires a thorough
knowledge of quantum electrodynamics and Maxwell’s electromagnetic field equations.
Similarly, a complete understanding of how we perceive the light our eyes see delves deeply into
the physiology and psychology of the human visual system. There is an enormous body of
literature related to the physical aspects of light as electromagnetic radiation (e.g., Hecht and
Zajac 1987) and an equally enormous amount devoted to how we perceive it (e.g., Cornsweet
1977). Fortunately, our interests are extremely modest. We simply want to measure what we see
and perceive.
3.
Radiometry
Radiometry is the science of measuring light in any portion of the electromagnetic spectrum. In
practice, the term is usually limited to the measurement of infrared, visible, and ultraviolet light
using optical instruments.
There are two aspects of radiometry: theory and practice. The practice involves the scientific
instruments and materials used in measuring light, including radiation thermocouples,
bolometers, photodiodes, photosensitive dyes and emulsions, vacuum phototubes, charge-coupled
devices, and a plethora of others. What we are interested in, however, is the theory.
Radiometric theory is such a simple topic that most texts on physics and optics discuss it in a few
paragraphs. Unfortunately, they rarely discuss the topic in enough detail to be useful. Shorn of
convoluted mathematics and obtuse definitions, radiometric theory is simple and easily
understood.
3.1
Radiant Energy
Light is radiant energy. Electromagnetic radiation (which can be considered both a wave and a
particle, depending on how you measure it) transports energy through space. When light is
absorbed by a physical object, its energy is converted into some other form. A microwave oven,
for example, heats a glass of water when its microwave radiation is absorbed by the water
molecules. The radiant energy of the microwaves is converted into thermal energy (heat).
Similarly, visible light causes an electric current to flow in a photographic light meter when its
radiant energy is transferred to the electrons as kinetic energy.
Radiant energy (denoted as Q) is measured in joules.
3.1.1 Spectral Radiant Energy
A broadband source such as the Sun emits electromagnetic radiation throughout most of the
electromagnetic spectrum, from radio waves to gamma rays. However, most of its radiant energy
is concentrated within the visible portion of the spectrum. A single-wavelength laser, on the other
hand, is a monochromatic source; all of its radiant energy is emitted at one specific wavelength.
From this, we can define spectral radiant energy, which is the amount of radiant energy per unit
wavelength interval at wavelength λ. It is defined as:
Qλ = dQ dλ
(1)
Spectral radiant energy is measured in joules per nanometer.
A-2
3.2
Radiant Flux (Radiant Power)
Energy per unit time is power, which we measure in joules per second, or watts. A laser beam, for
example, has so many milliwatts or watts of radiant power. Light “flows” through space, and so
radiant power is more commonly referred to as the “time rate of flow of radiant energy,” or
radiant flux. It is defined as:
Φ = dQ dt
(2)
where Q is radiant energy and t is time.
In terms of a photographic light meter measuring visible light, the instantaneous magnitude of the
electric current is directly proportional to the radiant flux. The total amount of current measured
over a period of time is directly proportional to the radiant energy absorbed by the light meter
during that time. This is how a photographic flash meter works -- it measures the total amount of
radiant energy received from a camera flash.
The flow of light through space is often represented by geometrical rays of light such as those
used in computer graphics ray tracing. They can be thought of as infinitesimally thin lines drawn
through space that indicate the direction of flow of radiant energy (light). They are also
mathematical abstractions -- even the thinnest laser beam has a finite cross section. Nevertheless,
they provide a useful aid to understanding radiometric theory.
Radiant flux is measured in watts.
3.2.1 Spectral Radiant Flux (Spectral Radiant Power)
Spectral radiant flux is radiant flux per unit wavelength interval at wavelength λ. It is defined as:
Φ λ = dΦ dλ
(3)
and is measured in watts per nanometer.
3.3
Radiant Flux Density (Irradiance and Radiant Exitance)
Radiant flux density is the radiant flux per unit area at a point on a surface, where the surface can
be real or imaginary (i.e., a mathematical plane). There are two possible conditions. The flux can
be arriving at the surface (Fig. 1a), in which case the radiant flux density is referred to as
irradiance. The flux can arrive from any direction above the surface, as indicated by the rays.
Irradiance is defined as:
E = dΦ dA
(4)
where Φ is the radiant flux arriving at the point and dA is the differential area surrounding the
point.
The flux can also be leaving the surface due to emission and/or reflection (Fig. 1b). The radiant
flux density is then referred to as radiant exitance. As with irradiance, the flux can leave in any
direction above the surface. The definition of radiant exitance is:
M = dΦ dA
(5)
where Φ is the radiant flux leaving the point and dA is the differential area surrounding the point.
A-3
dA
Figure 1a
dA
Irradiance
Figure 1b
Radiant exitance
The importance of a “real or imaginary” surface cannot be overstated. It means that radiant flux
density can be measured anywhere in three-dimensional space. This includes on the surface of
physical objects, in the space between them (e.g., in air or a vacuum), and inside transparent
media such as water and glass.
Radiant flux density is measured in watts per square meter.
3.3.1 Spectral Radiant Flux Density
Spectral radiant flux density is radiant flux per unit wavelength interval at wavelength λ. When
the radiant flux is arriving at the surface, it is called spectral irradiance, and is defined as:
Eλ = dE dλ
(6)
When the radiant flux is leaving the surface, it is called spectral radiant exitance, and is defined
as:
Mλ = dM dλ
(7)
Spectral radiant flux density is measured in watts per square meter per nanometer.
3.4
Radiance
Radiance is best understood by first visualizing it. Imagine ray of light arriving at or leaving a
point on a surface in a given direction. Radiance is simply the infinitesimal amount of radiant flux
contained in this ray. Period.
A more formal definition of radiance requires that we think of a ray as being an infinitesimally
narrow (“elemental”) cone with its apex at a point on a real or imaginary surface. This cone has a
differential solid angle dω that is measured in steradians.
We must also note that the ray is intersecting the surface at an angle. If the area of intersection
with the surface has a differential cross-sectional area dA, the cross-sectional area of the ray is
dA cos θ , where θ is the angle between the ray and the surface normal, as shown in Figure 2.
(The ray cross-sectional area dA cos θ is called the projected area of the ray-surface intersection
area dA. The same term is used when referring to finite areas ∆A.)
With these preliminaries in mind, we can imagine an elemental cone dω containing a ray of light
that is arriving at or leaving a surface (Figs. 3a and 3b). The definition of radiance is then:
L = d 2 Φ [dA(dω cos θ )]
(8)
A-4
n
Φ
θ
Projected area
dAcos θ
dA
Figure 2
A ray of light intersecting a surface
where Φ is the radiant flux, dA is the differential area surrounding the point, dω is the differential
solid angle of the elemental cone, and θ is the angle between the ray and the surface normal n at
the point.
Unlike radiant flux density, the definition of radiance does not distinguish between flux arriving
at or leaving a surface. In fact, the formal definition of radiance (ANSI/IES 1986) states that it
can be “leaving, passing through or arriving at” the surface.
Φ
n
Φ
n
θ
θ
dω
dω
dA
Figure 3a
Radiance (arriving)
dA
Figure 3b
Radiance (leaving)
Another way of looking a radiance is to note that the radiant flux density at a point on a surface
due to a single ray of light arriving (or leaving) at an angle θ to the surface normal is
dΦ (dA cos θ ) . The radiance at that point for the same angle is then d 2 Φ [dA(dω cos θ )] , or
radiant flux density per unit solid angle.
Radiance is measured in watts per square meter per steradian.
3.4.1 Spectral Radiance
Spectral radiance is radiance per unit wavelength interval at wavelength λ. It is defined as:
Lλ = d 3Φ [ dA(dω cos θ )dλ ]
(9)
and is measured in watts per square meter per steradian per nanometer.
A-5
3.5
Radiant Intensity
We can imagine an infinitesimally small point source of light that emits radiant flux in every
direction. The amount of radiant flux emitted in a given direction can be represented by a ray of
light contained in an elemental cone. This gives us the definition of radiant intensity:
I = dΦ dω I
(10)
where dω is the differential solid angle of the elemental cone containing the given direction. From
the definition of a differential solid angle ( dω = dA r 2 ), we get:
E = dΦ dA = dΦ r 2 dω = I r 2
(11)
where the differential surface area dA is on the surface of a sphere centered on and at a distance r
from the source and E is the irradiance of that surface. More generally, the radiant flux will
intercept dA at an angle θ (Fig. 4). This gives us the inverse square law for point sources:
E = I cos θ d 2
(12)
where I is the intensity of the source in the given direction and d is the distance from the source to
the surface element dA.
n
θ
d
dA
Figure 4
Inverse square law for point sources
We can further imagine a real or imaginary surface as being a continuum of point sources, where
each source occupies a differential area dA (Fig. 5). Viewed at an angle θ from the surface normal
n, the source has a projected area of dA cos θ . Combining the definitions of radiance (Eqn. 8) and
radiant intensity (Eqn. 10) gives us an alternative definition of radiance:
L = dI (dA cos θ )
(13)
where dI is the differential intensity of the point source in the given direction.
Radiant intensity is measured in watts per steradian.
n
dI
θ
dA
Figure 5
Radiance of a point source
A-6
3.4.1 Spectral Radiant Intensity
Spectral radiant intensity is radiant intensity per unit wavelength interval at wavelength λ. It is
defined as:
Iλ = dI dλ
(14)
and is measured in watts per steradian per nanometer.
4.
Illumination versus Thermal Engineering
The definitions given above are those commonly used in illumination engineering and are in
accordance with the American National Standard Institute publication Nomenclature and
Definitions for Illuminating Engineering (ANSI/IES 1986). Unfortunately, these definitions differ
somewhat from those used in thermal engineering (e.g., Siegel and Howell 1981). Radiative heattransfer theory (i.e., infrared light) does not use the point source concept that is common in
illumination engineering texts. Thermal engineers instead use the term “radiant intensity” to
describe radiance (watts per unit area per unit solid angle).
The different terminology was of little consequence until the computer graphics community
adapted the concepts of radiative heat transfer to create radiosity theory. In the process of doing
so it adopted thermal engineering’s terminology. This is an unfortunate situation, because
computer graphics also relies on the point source concept for ray tracing.
This tutorial defines radiant intensity as “watts per unit solid angle” and radiance as “watts per
unit area per unit solid angle” to maintain consistency between radiosity and ray-tracing theory.
Note, however, that many papers and texts on radiosity theory and some computer graphics texts
define “radiant intensity” as “watts per unit area per unit solid angle.”
5.
Photometry
Photometry is the science of measuring visible light in units that are weighted according to the
sensitivity of the human eye. It is a quantitative science based on a statistical model of the human
visual response to light -- that is, our perception of light -- under carefully controlled conditions.
The human visual system is a marvelously complex and highly nonlinear detector of
electromagnetic radiation with wavelengths ranging from 380 to 770 nanometers (nm). We see
light of different wavelengths as a continuum of colors ranging through the visible spectrum: 650
nm is red, 540 nm is green, 450 nm is blue, and so on.
The sensitivity of the human eye to light varies with wavelength. A light source with a radiance
of one watt/m2-steradian of green light, for example, appears much brighter than the same source
with a radiance of one watt/m2-steradian of red or blue light. In photometry, we do not measure
watts of radiant energy. Rather, we attempt to measure the subjective impression produced by
stimulating the human eye-brain visual system with radiant energy.
This task is complicated immensely by the eye’s nonlinear response to light. It varies not only
with wavelength but also with the amount of radiant flux, whether the light is constant or
flickering, the spatial complexity of the scene being perceived, the adaptation of the iris and
retina, the psychological and physiological state of the observer, and a host of other variables.
Nevertheless, the subjective impression of seeing can be quantified for “normal” viewing
conditions. In 1924, the Commission Internationale d’Eclairage (International Commission on
Illumination, or CIE) asked over one hundred observers to visually match the “brightness” of
monochromatic light sources with different wavelengths under controlled conditions. The
A-7
statistical result -- the so-called CIE photometric curve shown in Figure 6 -- shows the photopic
luminous efficiency of the human visual system as a function of wavelength. It provides a
weighting function that can be used to convert radiometric into photometric measurements.
1
0.9
0.8
0.7
Photopic 0.6
luminous 0.5
efficiency 0.4
0.3
0.2
0.1
0
390
440
490
540
590
640
690
740
Wavelength (nm)
Figure 6
CIE photometric curve
Photometric theory does not address how we perceive colors. The light being measured can be
monochromatic or a combination or continuum of wavelengths; the eye’s response is determined
by the CIE weighting function. This underlines a crucial point: The only difference between
radiometric and photometric theory is in their units of measurement. With this thought firmly in
mind, we can quickly review the fundamental concepts of photometry.
5.1
Luminous Intensity
The foundations of photometry were laid in 1729 by Pierre Bouguer. In his L’Essai d’Optique,
Bouguer discussed photometric principles in terms of the convenient light source of his time: a
wax candle. This became the basis of the point source concept in photometric theory.
Wax candles were used as national light source standards in the 18th and 19th centuries. England,
for example, used spermaceti (a wax derived from sperm whale oil). These were replaced in 1909
by an international standard based on a group of carbon filament vacuum lamps and again in
1948 by a crucible containing liquid platinum at its freezing point. Today the international
standard is a theoretical point source that has a luminous intensity of one candela (the Latin word
for “candle”). It emits monochromatic radiation with a frequency of 540 x 1012 Hertz (or
approximately 555 nm, corresponding with the wavelength of maximum photopic luminous
efficiency) and has a radiant intensity (in the direction of measurement) of 1/683 watts per
steradian.
Together with the CIE photometric curve, the candela provides the weighting factor needed to
convert between radiometric and photometric measurements. Consider, for example, a
monochromatic point source with a wavelength of 510 nm and a radiant intensity of 1/683 watts
per steradian. The photopic luminous efficiency at 510 nm is 0.503. The source therefore has a
luminous intensity of 0.503 candela.
5.2
Luminous Flux (Luminous Power)
Luminous flux is photometrically weighted radiant flux (power). Its unit of measurement is the
lumen, defined as 1/683 watts of radiant power at a frequency of 540 x 1012 Hertz. As with
luminous intensity, the luminous flux of light with other wavelengths can be calculated using the
CIE photometric curve.
A-8
A point source having a uniform (isotropic) luminous intensity of one candela in all directions
(i.e., a uniform intensity distribution) emits one lumen of luminous flux per unit solid angle
(steradian).
5.3
Luminous Energy
Luminous energy is photometrically weighted radiant energy. It is measured in lumen seconds.
5.4
Luminous Flux Density (Illuminance and Luminous Exitance)
Luminous flux density is photometrically weighted radiant flux density. Illuminance is the
photometric equivalent of irradiance, whereas luminous exitance is the photometric equivalent of
radiant exitance.
Luminous flux density is measured in lumens per square meter. (A footcandle is one lumen per
square foot.)
5.5
Luminance
Luminance is photometrically weighted radiance. In terms of visual perception, we perceive
luminance. It is an approximate measure of how “bright” a surface appears when we view it from
a given direction. Luminance used to be called “photometric brightness.” This term is no longer
used in illumination engineering because the subjective sensation of visual brightness is
influenced by many other physical, physiological, and psychological factors.
Luminance is measured in lumens per square meter per steradian.
6.
Lambertian Surfaces
A Lambertian surface is a surface that has a constant radiance or luminance that is independent of
the viewing direction. In accordance with the definition of radiance (luminance), the radiant
(luminous) flux may be emitted, transmitted, and/or reflected by the surface.
A Lambertian surface is also referred to as an ideal diffuse emitter or reflector. In practice there
are no true Lambertian surfaces. Most matte surfaces approximate an ideal diffuse reflector but
typically exhibit semispecular reflection characteristics at oblique viewing angles. Nevertheless,
the Lambertian surface concept is useful in computer graphics.
Lambertian surfaces are unique in that they reflect incident flux in a completely diffuse manner
(Fig. 7). It does not matter what the angle of incidence θ of an incoming geometrical ray is -- the
distribution of light leaving the surface remains unchanged.
We can imagine a differential area dA of a Lambertian surface. Being infinitesimally small, it is
equivalent to a point source, and so the flux leaving the surface can be modeled as geometrical
rays. The intensity Iθ of each ray leaving the surface at an angle θ from the surface normal is
given by Lambert’s cosine law:
Iθ = In cos θ
(15)
where Ιn is the intensity of the ray leaving in a direction perpendicular to the surface.
A-9
n
θ
dA
Figure 7
Reflection from a Lambertian surface
The derivation of Equation 15 becomes clear when we remember that we are viewing dA from an
angle θ. For a differential area dA with a constant radiance or luminance, its intensity must vary
in accordance with its projected area, which is dA cos θ . This gives us:
L = dI (dA cos θ ) = dIn dA
(16)
for any Lambertian surface.
There is a very simple relation between radiant (luminous) exitance and radiance (luminance) for
flux leaving a Lambertian surface:
M = πL
(17)
where the factor of π is a source of endless confusion to students of radiometry and photometry.
Fortunately, there is an intuitive explanation. Suppose we place a differential Lambertian emitter
dA on the inside surface of an imaginary sphere S (Fig. 8). The inverse square law (Eqn. 12)
provides the irradiance E at any point P on the inside surface of the sphere. However,
d = D cos θ , where D is the diameter of the sphere. Thus:
E = Iθ cos θ ( D cos θ ) = Iθ D2 cos θ
2
(18)
n
Sphere S
P
D
θ
θ
d
dA
Figure 8
A Lambertian emitter illuminating the interior of a sphere
and from Lambert’s cosine law (Eqn. 15), we have:
E = Iθ cos θ D2 cos θ = In D2
(19)
A-10
which simply says that the irradiance (radiant flux density) of any point P on the inside surface of
S is a constant.
This is interesting. From the definition of irradiance (Eqn. 4), we know that Φ = EA for constant
flux density across a finite surface area A. Since the area A of the surface of a sphere with radius r
is given by:
A = 4πr 2 = πD2
(20)
we have:
Φ = EA = πIn D2 D2 = πIn
(21)
Given the definition of radiant exitance (Eqn. 5) and radiance for a Lambertian surface (Eqn. 16),
we have:
M = dΦ dA = πdIn dA = πL
(22)
This explains, clearly and without resorting to integral calculus, where the factor of π comes
from.
7.
What is Radiosity?
ANSI/IES (1986) does not define or even mention the term “radiosity.” This is not unusual -there are many photometric and radiometric terms whose use is no longer encouraged.
“Illuminance,” for example, used to be called “illumination.” It was changed to “illuminance” to
avoid confusion with “the act of illuminating or the state of being illuminated” (ANSI/IES 1986).
When Moon wrote The Scientific Basis of Illumination Engineering in 1936, luminous exitance
was called “luminosity.” Curiously, there was no equivalent term for “radiant exitance,” so he
coined the term “radiosity” to describe the density of radiant flux leaving a surface.
The illumination engineering community ignored Moon’s proposal. Luminosity was changed to
“luminous emittance” and later to “luminous exitance,” with “radiant exitance” following as a
consequent. Meanwhile, the thermal engineering community adopted “radiosity” (e.g., Siegel and
Howell 1981).
It’s all very confusing. Fortunately, we only need to remember that:
Radiosity is radiant exitance.
This tutorial takes exception, perhaps unwisely, to the computer graphics community’s use of the
term “radiosity” to describe radiant exitance. Whereas it is an accepted term within the thermal
engineering community, it is not acceptable to illumination engineers for a variety of historical
reasons. The computer graphics and illumination engineering communities have many common
interests. If we are to communicate effectively, we must use a common lexicon of definitions.
That lexicon is ANSI/IES (1986).
8.
Conclusions
There is much more that we have not covered here, such as reflectance, transmittance, absorption,
scattering, diffraction, and polarization. We have also ignored the interaction of the human visual
system with light, including scoptic and mesopic luminous efficiency, temporal effects such as
flicker, and most important, color perception. The study of light and how we perceive it fills
volumes of research papers and textbooks.
A-11
Nevertheless, it begins with an understanding of how we measure light. While there is more to it
than “intensity,” you can see from this tutorial that radiometric and photometric theory is
conceptually simple and easily understood.
9.
References
ANSI/IES. 1986. Nomenclature and Definitions for Illuminating Engineering, ANSI/IES RP-161986. New York, NY: Illuminating Engineering Society of North America.
Bouguer, P. 1729. L’Essai d’Optique. Paris.
Cornsweet, T. N. 1977. Visual Perception. Cambridge, MA: Academic Press.
Hecht, E. and A. Zajac. 1987. Optics (2nd ed.) Reading, MA: Addison-Wesley.
Kaku, M. 1994. Hyperspace. Oxford, UK: Oxford University Press.
Moon, P. 1936. The Scientific Basis of Illumination Engineering. New York, NY: McGraw-Hill.
Siegel, R. and J. R. Howell. 1981. Thermal Radiation Heat Transfer. Washington, DC:
Hemisphere Publishing.
A-12
Download